when people say, “you can just do things” and “with AI, anyone can build anything instantly”, what they’re doing is selling you hype.

now, they’re not wrong, either, by the way. they’re making a good point. AI is a paradigm-shifting tool that opens a lot of previous locked doors for a lot of people. this ultimately means more opportunity, for those savvy enough to recognize it.

but you still need technical knowledge.

this is what they don’t tell you. without technical knowledge, AI is basically a search engine on steroids without the promoted results (yet, but it’s coming; just wait).

technical knowledge is what turns a good prompt into a great prompt. in a way, it highlights the importance of being articulate, and the age-old art of asking the right questions. you can get a lot more bang for your buck when prompting the more you specify what it is you want.

and how do you specify? well, it’s ultimately about visualization. the best results I get from AI is when I clearly visualize and articulate, in great detail, what exactly it is I’m looking for.

for example, you could ask for “the GDP of Argentina”. Ok, like, sure, you’ll get some information, and it’ll probably be fine. but it will also be broad, and possibly of weak sourcing. instead, you could ask for “the GDP of Argentina, with a table containing top 10 import and top 10 export partners with information on what sort of Things are being imported and exported”. the specificity of information in the second prompt yields much more insight.

(this isn’t how I would actually write a prompt; it’s just being used to highlight an example here)

always remember:

a system is what it does. a tool is what it’s used for.

the bottleneck in utility has always been the hand, not the hammer.